LeranMatch is part of the Visual Object Tracking Repository, which aims at providing a central repository for state-of-the-art tracking algorithms that are freely available. The source code for this tracker was obtained from its project website and extended by a challenge mode. The following description was copied literally from the original author.
Efficient Online Structured Output Learning for Keypoint-Based Object Tracking
Code to accompany the paper: Efficient Online Structured Output Learning for Keypoint-Based Object Tracking Sam Hare, Amir Saffari, Philip H. S. Torr Computer Vision and Pattern Recognition (CVPR), 2012
Copyright (C) 2012 Sam Hare, Oxford Brookes University, Oxford, UK
Contact: Sam Hare [email protected]
OpenCV: http://opencv.willowgarage.com/ Eigen: http://eigen.tuxfamily.org/
This code has been developed and tested using OpenCV v2.3.1 and Eigen v3.0.1
When running experiments, be sure to compile in Release mode, as Debug mode will be very slow.
learnmatch [--config config-file-path]
If no path is given the application will attempt to use ./config.txt.
Please see config.txt for configuration options.
The sequences used in the paper are available to download here:
http://www.samhare.net/research/keypoints
This code makes use of the following 3rd-party code:
OpenCV graphing utilities: http://www.shervinemami.co.cc/graphs.html BRISK: http://www.asl.ethz.ch/people/lestefan/personal/BRISK Online Boosting: http://www.vision.ee.ethz.ch/boostingTrackers/onlineBoosting.htm